INSTITUTION OF GAS ENGINEERS AND MANAGERS. Founded 1863 IGEM/TD/2 Edition 2 Royal Charter 1929 Communication XXXX Her Majesty the Queen

Transcription

1 INSTITUTION OF GAS ENGINEERS AND MANAGERS IGEM/TSP/12/156 Founded 1863 IGEM/TD/2 Edition 2 Royal Charter 1929 Communication XXXX Patron Her Majesty the Queen ASSESSING THE RISKS FROM HIGH PRESSURE NATURAL GAS PIPELINES DRAFT FOR COMMENT 1 This draft Standard IGEM/TD/2 Edition 2 has been prepared by a Panel under the chairmanship of Jane Haswell. 2 This Draft for Comment is presented to Industry for comments which are required by 18 th September 2012, and in accordance with the attached Reply Form. 3 This is a draft document and should not be regarded or used as a fully approved and published Standard. It is anticipated that amendments will be made prior to publication. It should be noted that this draft Standard contains intellectual property belonging to IGEM. Unauthorised copying or use by any unauthorised person or party is not permitted. 4 This is a copyright document of the Institution of Gas Engineers and Managers. Enquiries should be addressed in the first instance to: Nick Cowling IGEM IGEM House High Street Kegworth Derbyshire, DE74 2DA Tel: Fax: nick@igem.org.uk

2 The Organisations to which this Draft has been circulated are: AIGT BSi/GSE/33 DNO Collaboration Forum Energy Institute ENA EUA EUSkills Gas Forum GIRSAP GISG HSE Lloyds Register National Grid Northern Gas Networks Ofgem Scotia Gas Networks UKOPA Wales and West Utilities IGEM Committees Technical Enquiry Contacts

3 IGEM/TD/2 Edition 2 Communication XXXX Assessing the risks from high pressure natural gas pipelines Draft for Comment Founded 1863 Royal Charter 1929 Patron: Her Majesty the Queen

4 IGEM/TD/2 Edition 2 Communication XXXX Assessing the risks from high pressure natural gas pipelines Draft for Comment Price Code: C4H The Institution of Gas Engineers and Managers IGEM House High Street Kegworth Derbyshire, DE74 2DA Tel: Fax: general@igem.org.uk

5 Copyright 2012, IGEM. All rights reserved Registered charity number All content in this publication is, unless stated otherwise, the property of IGEM. Copyright laws protect this publication. Reproduction or retransmission in whole or in part, in any manner, without the prior written consent of the copyright holder, is a violation of copyright law. ISBN XX X ISSN Published by the Institution of Gas Engineers and Managers Previous Publications: Communication 1737 (2008) 1 st Edition For information on other IGEM Standards, visit our website,

6 CONTENTS SECTION PAGE 1 Introduction 1 2 Scope 5 3 Risk assessment of buried pipelines - Overview 6 4 Failure of a hazardous gas pipeline General Failure of a Natural Gas pipeline Stages of risk assessment Prediction of failure frequency Prediction of consequences Probability of ignition Thermal radiation and effects 14 5 Calculation of individual risk 16 6 Calculation of Societal risk 19 7 Multiple pipelines 23 8 Factors affecting risk level Damage mechanisms Factors for reduction of the external interference failure frequency for use in site-specific assessments Implementation of risk mitigation measures 30 APPENDIX 1 Glossary, acronyms, abbreviations, symbols, units and subscripts 33 2 References 35 3 Summary of HSE methodology for the provision of advice on planning developments in the vicinity of major accident hazard pipelines in the UK 38 4 Failure frequencies for UK pipelines 43 5 Examples of site-specific risk calculations 50 6 Recommended benchmark solutions for gas pipelines 54 FIGURE 1 Overview of contents of the Standard 4 2 Event tree for a Natural Gas pipeline failure 9 3 The stages of pipeline risk assessment 11

7 4 Calculation of pipeline length which can affect an individual at various distances from a pipeline 17 5 HSE framework for tolerability of risk 17 6 F-N criterion based on extensive application of IGEM/TD/ Site-Specific Pipeline Interaction Distance for Societal Risk 21 8 Reduction in external interference total failure frequency due to design factor 27 9 Reduction in external interference total failure frequency due to wall thickness Reduction in external interference total failure frequency due to depth of cover Reduction in external interference total failure frequency due to surveillance frequency Methodology for the prediction of site-specific risk levels for LUP developments Generic failure frequency curve for estimation of failure frequency due to external interference Proportion of ruptures to be applied to total failure rate derived from Figure Proposed development example Societal risk FN curves for proposed development 53 TABLE 1 Range of applicability of design factor and wall thickness reduction factors 28 2 Risk reduction factors for additional measures 30 3 Probability of barrier failure 38 4 Examples - Calculation of pipeline failure frequency due to external interference 45 5 Critical defect lengths and equivalent hole diameters for UKOPA pipeline cases operating at design factor R df = Failure frequency due to external corrosion vs wall thickness 47 7 Failure frequency due to material and construction defects vs wall thickness 48 8 Pipeline failure frequencies per 1000 km yr due to natural landsliding 49 9 Benchmark solutions 54

8 SECTION 1 : INTRODUCTION 1.1 This Standard revises and supersedes IGEM/TD/2, Communication 1737 which is obsolete. 1.2 This Standard has been drafted by a Panel appointed by the Institution of Gas Engineers and Managers (IGEM) Gas Transmission and Distribution Committee, subsequently approved by that Committee and published by the authority of the Council of IGEM. 1.3 The Standard has been updated to more clearly differentiate between the assessments that would be undertaken by pipeline operators to justify the safe operation of the pipeline and assessments and those undertaken for land use planning purposes. Additionally, some of the technical information within the Standard, including risk reduction factors for concrete slab protection over pipelines and the risk reduction factors assumed for increased depth of cover over the pipeline have also been updated. 1.4 This Standard makes use of the terms must, shall and should when prescribing particular requirements. Notwithstanding Sub-Section 1.9: the term must identifies a requirement by law in Great Britain (GB) at the time of publication the term shall prescribes a requirement which, it is intended, will be complied with in full and without deviation the term should prescribes a requirement which, it is intended, will be complied with unless, after prior consideration, deviation is considered to be acceptable. 1.5 IGEM/TD/1 Edition 5, Section 4 on planning and legal considerations, provides guidance on the route selection and location of new pipelines in various areas in terms of the acceptable proximity to significant inhabited areas. IGEM/TD/1 Edition 5, Section 6 on design, categorizes locations adjacent to pipelines into Type R, S and T according to population density and/or nature of the immediate surrounding area. IGEM/TD/1 Edition 5, Section 12 on operations and maintenance, provides requirements for surveillance and inspection which will reveal encroachment into areas of interest adjacent to a pipeline. Significant developments or infringements may require risk assessment using societal risk analysis for comparison with suitable risk criteria to allow the operator to assess whether the risks remain within acceptable limits. IGEM/TD/1 Edition 5, Appendix 3 on risk assessment techniques, describes the application of risk assessment and includes a description of societal risk assessment with a sample of an actual F-N criterion based upon extensive application of previous editions of IGEM/TD/1. IGEM/TD/2 aims to support pipeline operators when carrying out risk assessments to assess safety risks associated with planning developments in close proximity to pipelines. 1.6 The general approach to the risk assessment process follows the stages outlined in IGEM/TD/1 Edition 5 Appendix 3. IGEM/TD/2 includes guidance on: determining failure frequencies consequence modelling standard assumptions to be applied in the risk assessment methodology conducting site-specific risk assessments risk reduction factors to be applied for mitigation methods benchmark results for individual and societal risk levels. 1

9 This Standard provides guidance for the risk assessment of major hazard pipelines containing Natural Gas. The need for undertaking a pipeline risk assessment may typically arise as a result of the need to: assess hazards and risks in support of the pipeline operator s Major Accident Prevention Document (MAPD); assess the acceptability of a development or developments that do not comply with proximity requirements or the population density requirements of IGEM/TD/1; support operational changes to a pipeline e.g. uprating (increasing the operating pressure) of a pipeline; assess the risks associated with specific operational issues; assess the implications of a Land Use Planning Application (see below). Under the Town and Country Planning Act in England and Wales, and the Town and Country Planning Act (Scotland) in Scotland, it is the Local Planning Authority s responsibility to determine the acceptability of individual planning applications including developments in the vicinity of high pressure gas pipelines. These decisions would take account of safety advice provided by the Health and Safety Executive (HSE). In coming to a decision the Local Authority would weigh local needs and benefits and other planning considerations alongside the HSE s advice. The HSE s advice on land use planning in the vicinity of high pressure pipelines is delivered through PADHI (planning advice for developments near hazardous installations). A summary of the HSE s risk methodology upon which the PADHI advice is based is provided in Appendix 3 of this document. In the event of a Local Planning Authority determining that a planning application should not be allowed based on the HSE s advice, the developer may approach the pipeline operator or seek independent guidance on the measures that can be taken to further reduce the risk. Alternatively there may be pipeline risk reduction features at the location of the proposed development that were not fully taken into account by the Local Planning Authority or the HSE when applying PADHI. These could, for example, include sections of pipeline with wall thicknesses greater than the notified pipeline wall thickness. The approaches detailed in this document can be used to undertake further detailed quantitative risk assessments in relation to land use planning applications. As outlined in Reference 19, the HSE take a different approach when assessing the acceptability of a proposed development that has not yet received planning permission compared with an existing development. Not allowing the development is seen by the HSE as being relatively inexpensive when compared to the costs entailed in requiring existing developments with similar risks to introduce remedial measures. Pipeline operators and developers need to be aware of these differences in approach when undertaking assessments in relation to land use planning applications. Further details on the HSE s approach for assessing the acceptability of proposed developments in the vicinity of high pressure pipelines are provided in Appendix 3. The guidance in this document does not cover environmental risks. 1.7 An overview of this Standard s content is given in Figure 1. The guidance in this Standard is provided for the benefit of pipeline operators, local planning authorities, developers and any person involved in the risk assessment of developments in the vicinity of existing high pressure Natural Gas pipelines. It is based on the established best practice methodology for pipeline risk assessment, and is intended to be applied by competent risk assessment practitioners. 2

10 Where significant numbers of people are exposed to the risk, the pipeline operator may wish to carry out risk assessment using societal risk analysis for comparison with suitable risk criteria to allow the operator to assess whether the risks remain within acceptable limits. Section 6 describes the application of societal risk, and includes reference to the recommended F-N envelope in IGEM/TD/1 Edition It is now widely accepted that the majority of accidents in industry generally are in some measure attributable to human as well as technical factors, in the sense that actions by people initiated or contributed to the accidents, or people might have acted better to avert them. It is therefore necessary to give proper consideration to the management of these human factors and the control of risk. To assist in this, it is recommended that due cognisance should be taken of HSG48. The primary responsibility for compliance with legal duties rests with the employer. The fact that certain employees, for example responsible engineers, are allowed to exercise their professional judgement does not allow employers to abrogate their primary responsibilities. Employers must: (a) Have done everything to ensure, so far as is reasonably practicable, that there are no better protective measures that can be taken other than relying on the exercise of professional judgement by responsible engineers. (b) Have done everything to ensure, so far as is reasonably practicable, that responsible engineers have the skills, training, experience and personal qualities necessary for the proper exercise of professional judgement. (c) Have systems and procedures in place to ensure that the exercise of professional judgement by responsible engineers is subject to appropriate monitoring and review. (d) Not require responsible engineers to undertake tasks which would necessitate the exercise of professional judgement that is beyond their competence. There should be written procedures defining the extent to which responsible engineers can exercise their judgement. When responsible engineers are asked to undertake tasks that deviate from this, they should refer the matter for higher review. Note: The responsible engineer is a suitably qualified, competent and experienced engineer or a suitably qualified, competent and experienced person acting under his or her supervision, appointed to be responsible for the application of all or part of this Standard. 1.9 Notwithstanding Sub-Section 1.5, this Standard does not attempt to make the use of any method or specification obligatory against the judgement of the responsible engineer. Where new and better techniques are developed and proved, they should be adopted without waiting for modification to this Standard. Amendments to this Standard will be issued when necessary and their publication will be announced in the Journal of IGEM and other publications as appropriate Requests for interpretation of this Standard in relation to matters within its scope, but not precisely covered by the current text, should be addressed to Technical Services, IGEM, IGEM House, High Street, Kegworth, Leicestershire, DE74 2DA, and will be submitted to the relevant Committee for consideration and advice, but in the context that the final responsibility is that of the engineer concerned. If any advice is given by, or on behalf of, IGEM, this does not relieve the responsible engineer of any of his or her obligations As with any risk assessment, judgement has to be employed by the risk assessor at all stages of the assessment. IGEM/TD/2 is intended to support the application of expert judgement. The final responsibility for the risk assessment lies with the assessor, and it is essential that the assessor be able to justify 3

11 every key assumption made in the assessment and document these assumptions as part of the assessment. FIGURE 1 - OVERVIEW OF CONTENTS OF THE STANDARD 4

12 SECTION 2 : SCOPE 2.1 IGEM/TD/2 provides a framework for carrying out an assessment of the acute safety risks associated with major accident hazard pipelines (MAHPs) containing high pressure Natural Gas. It provides guidance on the selection of pipeline failure frequencies and on the modelling of failure consequences for the prediction of individual and societal risks. The principles of this Standard are based on best practice for the quantified risk analysis of new pipelines and existing pipelines. It is not intended to replace or duplicate existing risk analysis methodology, but is intended to support the application of the methodology and provide guidance on its use. 2.2 This Standard is applicable to buried pipelines on land that can be used to carry high pressure Natural Gas, that is hazardous by nature, and therefore liable to cause harm to persons. It is limited to cross country pipelines and is not intended for application to pipelines and pipework forming part of above-ground installations, nor to associated equipment such as valves. The Standard does not cover environmental risks. 2.3 This Standard is intended for use in assessing the risks from high pressure gas pipelines including the additional risks that arise as a result of new developments in the vicinity of pipelines. This Standard provides a framework to help inform the pipeline operator on the acceptability, or otherwise, of these risks. 2.4 All references to gas pressure are gauge pressure, unless otherwise stated. 2.5 Details of all standards and other publications referenced are provided in Appendix 2. Where standards are quoted, equivalent national and international standards, etc. equally may be appropriate. 2.6 Italicised text is informative and does not represent formal requirements. 2.7 Appendices are informative and do not represent formal requirements unless specifically referenced in the main sections via the prescriptive terms must, shall or should. 5

13 SECTION 3 : RISK ASSESSMENT OF BURIED PIPELINES - OVERVIEW 3.1 The failure of a pipeline containing Natural Gas has the potential to cause serious damage to the surrounding population, property and the environment. Failure can occur due to a range of potential causes, including accidental damage, corrosion, fatigue and ground movement. The safety consequences of such a failure are primarily due to the thermal radiation from an ignited release. Quantified risk assessment (QRA) applied to a pipeline involves the calculation of risk resulting from the frequencies and consequences of a complete and representative set of credible accident scenarios. In general terms, QRA of a hazardous Natural Gas pipeline consists of the following stages: (a) Gathering data (pipeline and its location, meteorological conditions, physical properties of the substance, population) (Sub-Section 4.3). (b) Prediction of the frequency of the failures to be considered in the assessment (Sub-Section 4.4). (c) Prediction of the consequences for the various failure scenarios (Sub- Sections 4.5, 4.6, 4.7), including: calculation of release flow rate determination of ignition probability calculation of the thermal radiation emitted by fire in an ignited release quantification of the effects of thermal radiation on the surrounding population. (d) Calculation of risks and assessment against criteria: estimation of individual risk (Section 5). estimation of societal risk (Section 6). (e) Consideration of multiple pipelines (Section 7). (f) Identification of site-specific risk reduction measures (Section 8). Pipeline failure frequency is usually expressed in failures per kilometre year or per 1000 kilometre years. Failure frequency should be predicted using verified failure models and predictive methodologies (A2.3 Refs 1, 2, 3 and 4) or otherwise derived from historical incidents that have occurred in large populations of existing pipelines that are representative of the population under consideration, as recorded in recognised, published pipeline data (A2.3 Refs 5 and 6). Various factors may then be taken into account for the specific pipeline design and operating conditions to obtain the failure rate to be applied. Note: Predictive models can be generated for all damage types and failure modes depending on the data available. In the UK, external interference is the dominant mode, and predictive models based on operational data are available (A2.3 Refs 1, 2 and 3). In general, failure frequency due to other damage types is typically derived using historical data (A2.3 Refs 5 and 6). 3.2 The consequences of pipeline failures should be predicted using verified mathematical models, the results validated using experimental data at various scales up to full or by comparison with recognised solutions, as well as comparison of model predictions with the recorded consequences of real incidents. The results of a consequence analysis should take into account all feasible events, in terms of the effective distance over which people are likely to become casualties. This should take into account people both outdoors and indoors. 6

14 3.3 Pipelines present an extended source of hazard, and can pose a risk to developments at different locations along their route. Where a length of pipeline over which a location-specific accident scenario can affect the population associated with a specific development, the full length over which a pipeline failure could affect the population or part of the population should be considered in the risk assessment. This length is known as interaction distance (see Sections 5 and 6). 7

15 SECTION 4 : FAILURE OF A HAZARDOUS GAS PIPELINE 4.1 GENERAL Failure of a hazardous Natural Gas pipeline has the potential to cause damage to the surrounding population and property. Failure can occur due to a range of potential causes, including accidental damage, corrosion, fatigue and ground movement. The consequences of failure are primarily due to the thermal radiation that is produced if the release ignites. 4.2 FAILURE OF A NATURAL GAS PIPELINE Failure of a high pressure Natural Gas pipeline is a leak or rupture caused by damage such as external interference, corrosion, fatigue or ground movement. Leaks are defined as gas lost through a stable defect; ruptures are defined as gas lost through an unstable defect which extends during failure to result in a full break or failure of an equivalent size in the pipeline. The escaping gas may ignite, resulting in a fireball and/or a crater fire or jet fire which generates thermal radiation. An event tree for the failure of a Natural Gas pipeline is shown in Figure If immediate ignition of a rupture release of gas occurs, a fireball can be produced which typically lasts for up to 30 s and is followed by a crater fire. If ignition is delayed by 30 s or more, it is assumed that only a crater fire will occur. For the assessment of a rupture release, it is normally assumed that the ends of the failed pipe remain aligned in the crater and the jets of gas interact. However, under specific conditions, for example at a location close to a bend, it is possible for one or both pipe ends to become misaligned and produce one or two jets which are directed out of the crater and are unobstructed. Such releases can produce directional effects making their assessment more complex. Where such a location or pipe is being assessed, the standard case would normally be assessed and then the sensitivity of the location to directional releases reviewed. A more detailed assessment may then be required which would go beyond the standard methodology described in this Standard. Note: For pipelines of diameter exceeding 300 mm, the standard assumption is that the pipe ends are aligned For high pressure Natural Gas pipelines, the release has a large momentum flux at the source and this normally has a significant vertical component. Natural Gas is lighter than air and so, over the normal range of transmission pipeline pressures, the jet or plume leaving the crater can be assumed to be buoyant. Hence, the possibility of a release of Natural Gas leading to a flammable mixture at ground level and a potential flash fire hazard affecting developments in the vicinity of a pipeline is highly unlikely. 8

16 FIGURE 2 - EVENT TREE FOR A NATURAL GAS PIPELINE FAILURE 9

17 4.3 STAGES OF RISK ASSESSMENT The stages of pipeline risk assessment are represented in Figure 3. In general terms, a QRA of a Natural Gas pipeline should consist of 4 stages: (a) input of data (pipeline and its location, meteorological conditions, physical properties of gas, population) (b) prediction of failure mode and frequency for each credible failure cause (c) prediction of consequences, calculation of release flow rate determination of ignition probability calculation of thermal radiation emitted by fire in an ignited release quantification of the effects of thermal radiation on the surrounding population (d) calculation of risks The first stage of the risk assessment process is to gather the required data to characterise the pipeline, its contents and the surrounding environment. These data are used at various stages of the analysis. The data should be obtained from engineering records, operating data, the pipeline operating limits in the pipeline notification and from an examination of the pipeline surroundings. The principle input data required for a pipeline QRA are: pipeline geometry outside diameter; wall thickness pipeline material properties for example grade, specified minimum yield strength (SMYS), tensile strength (TS), toughness (or Charpy impact value) pipeline operational parameters maximum operating pressure (MOP), temperature, pipeline shut-down period and boundary conditions location details, including; length and route of the pipeline to be assessed area type (rural, suburban) depth of cover additional protection measures for the pipeline, for example concrete slabbing details of any above- and below-ground pipeline marking development and building categories in the vicinity and their distance from the pipeline population and occupancy levels within the consequence range of the pipeline road/rail crossing details, including traffic density river crossings physical properties of the gas being transported, including information to characterise the pressure, volume and temperature behaviour of the gas throughout the range of conditions relevant to the analysis atmospheric conditions and wind speed and direction. Any site-specific variations in the data should be assessed, and justifications for any additional assumptions to be applied locally should be documented. For depth of cover, site-specific depths should be taken into account. Where additional pipeline protection such as slabbing is to be taken into account, the design and installation should be assessed to ensure that additional loading is not imposed upon the pipeline, and direct contact should be maintained between the pipe coating and the surrounding soil. The stages of pipeline risk assessment are schematically represented in Figure 3. 10

18 Input Parameters Failure cause? External interference Corrosion Material and construction defects Ground movement Other causes e.g fatigue Causes Risk transect Calculation of Failure Frequency Risk Calculations Individual Societal FN curve (PLL, EV) Failure mode? Rupture or Puncture Outflow Dispersion Ignition Consequence calculations Thermal radiation Radiation effects FIGURE 3 - THE STAGES OF PIPELINE RISK ASSESSMENT 4.4 PREDICTION OF FAILURE FREQUENCY Failure of a pipeline can occur due to a number of different causes such as: external interference corrosion (internal and external, including stress corrosion cracking (SCC) and alternating current (AC)/direct current (DC) - induced corrosion) material or construction defects ground movement other causes, such as fatigue, operational errors etc. The failure modes which should be considered include leaks and line breaks or ruptures. Leak sizes range from pinholes up to hole sizes which represent critical or unstable defects for particular pipeline parameters. Unstable defects result in ruptures. A rupture release is a full bore, double-ended break or equivalent from which gas is released into a crater from both sections of pipe. Typical failure frequencies for UK pipelines are given in Appendix 4. Note: In most cases, the risk from natural gas pipelines will be dominated by the rupture scenario Leaks should be considered in terms of a specific range of hole sizes. Usually, failure frequency data is quoted for the sum of all hole sizes, and these should be classified into specific hole sizes to enable the risk assessment to be carried out. To determine the range of hole sizes to be considered in the consequence assessment, the hole size which gives an equivalent outflow to the critical length of an axial defect for specific pipeline parameters should be determined. 11

19 Typical hole diameters which are equivalent to critical defect lengths for high pressure Natural Gas pipelines (A2.3 Ref 7) are given in Appendix 4. Note 1: Critical defect length and equivalent hole diameter applies to external interference where axial, crack-like defects can occur; the equivalent hole sizes which relate to such defects do not apply to rounded punctures or stable holes due to corrosion or material and construction defects. Note 2: The maximum possible hole size in high pressure gas pipelines is limited according to the critical defect size. Typical failure frequencies for UK MAHPs are given in Appendix 4. Where other data sources are used, these should be documented In a risk assessment, the likelihood of each failure scenario should be evaluated and expressed in terms of failure frequency and pipeline unit length. The usual form is to express the failure rate in terms of failures per kilometre per year, or per 1000 kilometres per year (equivalent to failures per million metres per year). 4.5 PREDICTION OF CONSEQUENCES A consequence calculation should model and predict the transient gas release rate, the ignition probabilities, the characteristics of the resulting fire (fireball, and/or crater fire or jet fire), the thermal radiation field produced and the effects of the radiation on people and buildings nearby. Fires which should be considered as a result of ignition of a large gas release caused by a rupture are as follows: fireball, which occurs in the event of immediate ignition of a large gas release crater fire, which occurs in the event of delayed ignition of the gas flow released into the crater formed by the release, or following the immediate ignition fireball Relevant references are A2.3 Refs 4 and 8 to The following aspects should be considered: outflow as a function of time (influenced by failure location and upstream and downstream boundary conditions). Pipeline rupture outflow requires complex calculations involving pressure reduction in the pipe (A2.3 Refs 4, 10, 12 and 13). Outflow from holes is calculated using conventional sharpedged orifice equations for gas using a suitable discharge coefficient (A2.3 Ref 11) thermal radiation from the initial and reducing outflow into the fireball plume if the release is ignited immediately thermal radiation from jet and crater fires fed from reducing or steady state outflows. Other consequences, which are generally found to be negligible compared to fire effects, include: release of pressure energy from the initial fractured section pressure generated from combustion during the initial phase if the release is ignited immediately missiles generated from overlying soil or from pipe fragments. 12

20 4.5.3 The consequence model should also consider: wind speed (often taken to be 2 m s -1 at night and 5 m s -1 during the daytime) which affects the fire tilt and, hence, the resulting radiation effects wind direction only required for a site specific risk assessment where wind direction will affect the populated area being considered in a non-uniform way humidity this affects the proportion of thermal radiation absorbed by the atmosphere As observed in actual events and experimental research, if ignition occurs immediately on, or shortly after, a rupture, a transient fireball could occur. The fireball, which is the result of combustion of a mushroom-shaped cap that is fed from below by the established part of the fire, lasts typically for up to 30 seconds (depending on pipeline diameter and initial pressure) In modelling fires following a rupture, the transient nature of the release should be modelled. This calculation requires an estimate of the initial and steady state release rates and an estimate of the inventory of the pipeline network which is discharging to the release point. For generic calculations, an assumption is that the break occurs half-way between a compressor station (or pressure regulating installation (PRI)) and the downstream compressor check valve or pressure regulating installation, with pressure being maintained from the upstream compressor station and no reverse flow occurring at the compressor station check valve or pressure regulating installation In modelling jet fires from punctures, the release can be considered to be steady-state. Usually the consequence model considers a vertical jet flame, with wind tilt created by the current wind velocity. More elaborate models are possible with different angles of flame. The consequences predicted by such models are increased directionally but the conditional probability is reduced. 4.6 PROBABILITY OF IGNITION The risks from a pipeline containing flammable Natural Gas depend critically on whether a release is ignited, and whether ignition occurs immediately or is delayed. Generic values for ignition probability can be obtained from data from historical incidents. A trend has been observed from analysis of historical data for rupture incidents where the ignition probability increases linearly with pd 2 (A2.3 Ref 14). The correlation derived for rupture releases takes the form: P ign = pd 2 ; 0 pd 2 57 and P ign = 0.81; pd 2 > 57 P ign = probability of ignition p = pipeline operating pressure (bar) d = pipeline diameter for ruptures (m) The various ignition possibilities, together with the release types, are drawn out logically on an event tree (see Figure 2) to obtain overall probabilities (A2.3 Refs 10 and 14). Appropriate values for the probability of immediate or delayed ignition (and, if delayed, the assumed time(s) of ignition) should also be selected. The probability of ignition P ign calculated as detailed above is then generally apportioned as 0.5 for immediate ignition and 0.5 for delayed ignition, where delayed ignition occurs after 30 seconds. 13

21 For puncture releases, the same ignition probability relationship may be applied, with d equal to the release hole diameter and with the pd 2 value halved, reflecting the difference between the two sources contributing to the gas release following a rupture and the single source contributing to a puncture release. 4.7 THERMAL RADIATION AND EFFECTS Fatal injury effects are assumed for cases where people in the open air or in buildings are located within the flame envelope from a fireball, crater fire or jet fire. Outside the flame envelope, the effects are dependent on direct thermal radiation from the flame to the exposed person or building. For crater fires, thermal radiation is highest at the flame surface and is attenuated with increasing distance as the view factor reduces and as heat is absorbed into the atmosphere. Thermal radiation is calculated from the energy of the burning material. There are two main methods of calculation in use; the View Factor method which assumes a surface emissive power from the flame, and the Point Source method which assumes all the energy is emitted from one (or several) point sources within the flame. Usually the energy from the fireball pulse is calculated using the View Factor method. The unit of thermal radiation dose is then defined as: Thermal Dose Unit (tdu) = (W) 4/3 x t W = flux = intensity of thermal radiation (kw m -2 ) t = time (s). Note: W is not independent of time for a transient release and is, normally, summed over exposure until safe shelter, the dose limit or a cut-off thermal radiation level of 1 kw m -2 for example, is recorded Experimental and other data indicate that thermal radiation dose levels can have different effects within a population depending on individual tolerance. The variation of effects has been estimated from burn data for human beings which suggests that the radiation level causing 50% fatal injuries in an average population can be taken as 1800 tdu. This level of thermal dose is often used in risk assessments. Developments such as schools, hospitals and old peoples homes may be classed as sensitive developments due to the increased vulnerability of the population groups involved to harm from thermal radiation hazards. For such sensitive developments, where a more cautious approach may be appropriate, a casualty criterion of 1% lethality (corresponding to a dangerous dose ) should be applied as this corresponds to the 50% lethality level for this group of people. Such an approach allows sensitive developments to be included in an assessment of a wider mix of developments. Note 1: The dangerous dose concept is defined as causing: (a) severe distress to almost everyone in the area (b) a substantial fraction of the exposed population requiring medical attention (c) some people being seriously injured, requiring prolonged treatment (d) high-susceptible people being killed. Note 2: To assess safe escape distance, a number of factors need to be taken into account, including speed of escape for people outside running away from the fire, location and types of buildings, populations indoors, daytime and night, etc. Note 3: HSE apply the 1000 tdu criterion for dangerous dose when determining the distances to the boundaries of the zones used to assess the acceptability of proposed new developments in the vicinity of high pressure pipelines, this is described in Appendix 3. Note 4: For a more direct comparison with an assessment carried out by HSE for land use planning purposes, the assessment can be carried out using the dangerous dose level of thermal radiation for the whole population. Comparison with a societal risk acceptability criterion based on a higher level of thermal radiation will then be conservative for an average or mixed population. 14

22 4.7.4 The effects of the time-varying thermal radiation field generated by the fires (see sections and 4.5.5) are quantified by assessing the variation of the thermal radiation field with time, typically by assessing the radiation at a series of times following ignition or from an equivalent steady state interpretation of the event, calculating the resultant dose levels for selected time intervals, then summing the thermal radiation dose received during the time of exposure at a fixed point, or by a moving target, in order to predict the effects on buildings and on persons who can attempt to escape from the effects of the fire. The distance from the fire at which buildings would burn, and also the distance at which persons are able to escape from the effects of the fire, are predicted based on either threshold ignition criteria for buildings and casualty criterion method or a probit equation method for persons. In the casualty criterion method, the thermal radiation predictions are used to calculate the distances at which the specified criteria are met. The casualty criterion method does not directly account for the fact individuals within the population have a range of sensitivity to thermal radiation; however, this can be taken into account using a probit equation method. In this method, escape probability calculations are performed for each point on each escape path for each 1% band of the population in turn. These values are then integrated over the whole population to arrive at an overall casualty rate. In each method, the possibility that the escaping person may find shelter in a building beyond the building-burning distance can be accounted for. Relevant references are A2.3 Refs 10, 17 and

23 SECTION 5 : CALCULATION OF INDIVIDUAL RISK Individual risk is a measure of the frequency at which an individual at a specified distance from a pipeline is expected to sustain a specified level of harm from the realization of specific hazards. A simple explanation of the calculation of individual risk posed by pipelines is given in this section. Further detail may be found in the references listed in A2.3 (4,8,10,12). Risk calculations should be carried out by competent experts using appropriate, quality assured software. Individual risk contours for pipelines of given geometry, material properties and operating conditions form lines parallel to the pipeline axis. The distance from the pipeline at which a particular level of risk occurs depends upon the pipeline diameter, operating pressure, frequency of failure and failure mode. The risks from the various failure scenarios (ruptures and various holes sizes causing fireballs, crater fires and jet fires) should be collated and the individual risk profile at various distances plotted on a graph. From this plot, it is possible to identify the risk of a specific effect, for example fatality or dangerous dose, to an individual at a given distance from the pipeline. Shown in cross-section perpendicular to the pipeline, the risk levels are known as the risk transect. Pipelines present a hazard along the pipeline route and, therefore, the full length over which a pipeline failure could affect any specific location should be considered in the risk assessment. This length is known as the interaction distance. For a simple model where wind speed conditions are zero, the consequences are circular and the interaction distance is calculated as shown in Figure 4. The interaction distance shown can be multiplied by the pipeline failure frequency, the probability of ignition and the consequences (thermal radiation and casualty criteria) to obtain the risk at any distance from the point of release. Distances to specified Individual Risk levels can be obtained from the Risk Transect. 16

24 FIGURE 4 - CALCULATION OF PIPELINE LENGTH WHICH CAN AFFECT AN INDIVIDUAL AT VARIOUS DISTANCES FROM A PIPELINE Criteria for individual risk levels have been determined by HSE in the UK. The framework for the tolerability of individual risk published by HSE, based on historical risk of death (see A2.3 Ref 13), is shown in Figure 5. FIGURE 5 - HSE FRAMEWORK FOR TOLERABILITY OF RISK 17

25 HSE sets land use planning zones for major hazard sites, including high pressure pipelines transporting defined hazardous substances based on the individual risk levels for dangerous dose. Three risk-based zones, the inner, middle and outer zones, are defined by HSE based on the dangerous dose or worse. The outer zone is defined as the consultation distance within which the risk implications of planning developments should be considered by the Local Planning Authority (LPA). Land use planning zones applied to major accident hazard pipelines in the UK defined by HSE are discussed in Appendix 3. 18

26 SECTION 6 : CALCULATION OF SOCIETAL RISK 6.1 Societal risk is a measure of the relationship between the frequency of an incident and the number of casualties that will result. Societal risk can be generic, in which a constant distributed population in the vicinity of the pipeline is assumed, or site-specific, in which the details of particular developments, building layout and population distributions are taken into account. Site-specific assessments are needed for housing developments, industrial premises, workplaces such as call centres, commercial and leisure developments, and any development involving sensitive populations. 6.2 The hazards associated with high pressure Natural Gas pipelines tend to be high consequence, low frequency events, and societal risk is generally the determining measure for the acceptability of pipeline risk. The calculation is carried out by assessing the frequency and consequences of all of the various accident scenarios which could occur along a length of pipeline. 6.3 Societal risk is typically expressed graphically using an F-N curve showing the cumulative frequency F (usually per year) of accidents causing N or more casualties. For application to pipelines, it is necessary to specify a length over which the frequency and consequences of all accident scenarios are collated. Application of IGEM/TD/1 has been shown to result in a residual societal risk and this is expressed as a criterion F-N envelope in Figure 6. This criterion is calculated from the residual risk to a range of generic R area cases at the limited design factor and population density and for S area cases at the limiting design factor and for typical population densities. The envelope was drawn around the resultant family of approximately rectangular shaped F-N curves and was subjected to As Low As Reasonably Practicable (ALARP) considerations. In effect, it is a practical representation of the broadly acceptable limit of a typical code-compliant pipeline route. Note 1: This envelope has been generated using a methodology which incorporated thermal radiation dose of 1800 tdu (50% fatality) for the casualty criterion to assess population effects. Note 2: The methodology applied must be consistent with the criterion used in assessment of results. Note 3: The IGEM/TD/1 F-N criterion envelope represents broadly acceptable risk levels in both R and S Areas. 19

27 FIGURE 6 - F-N CRITERION BASED ON EXTENSIVE APPLICATION OF IGEM/TD/1 6.4 Societal risk is of particular significance to pipeline operators and the impact of multiple fatality accidents on people and society in general. The original routing of the pipeline is expected to have taken into account the population along the route, but infill and incremental developments may increase the population in some sections of the route. Societal risk assessment allows these developments to be assessed against the original routing criteria. When societal risk has increased significantly, the pipeline operator may need to consider justifiable mitigation measures to reduce risk. Because of the dominance of the rupture failure mode, F-N curves from typical R area assessments of minor code infringements and land use planning cases are fairly close to rectangular in shape. This will also apply to many S area assessments. However, some smaller diameter pipelines and/or complex S area 20

28 locations can produce F-N curves which differ from this approximately rectangular shape. Particular care will be required in assessing the results in these cases (see Sub-Section 6.7). 6.5 Usually, population density is not equally spread so, in some cases, clusters of population occur along a pipeline route. Assessment of the societal risk in accordance with the F-N envelope may still allow this to be classified as an acceptable situation not requiring any upgrading of the pipeline to reduce the risk. It is recommended that this form of assessment is carried out on an ongoing basis to ensure that the risks due to developments and changes occurring near the pipeline are monitored, quantified and managed. Note: Societal risk assessment provides an operator s view on the acceptability of a development in the vicinity of a pipeline, which takes account of other developments in the vicinity of the pipeline. 6.6 The methodology for considering and assessing risk scenarios, failure cases, failure frequencies and consequences is similar to that used to obtain individual risk levels. 6.7 To carry out a site-specific societal risk assessment, the maximum distance over which the worst case event could affect the population in the vicinity should be determined, for example the site length combined with the escape distance for thermal radiation (see Figure 6). This is defined as the site interaction distance. 6.8 The accident scenarios which are relevant for the pipeline section within the site interaction distance should be listed and the actual population density within the area of interest should be determined. The frequency, f, and effect area for each accident scenario is then assessed at points along the site interaction distance, and the number of people, N, who would be affected, is determined at a specific location. This provides a number of f-n pairs, which are then ordered with respect to increasing number of casualties, N, and the cumulative frequency, F of N of more people being affected is determined, giving a site-specific F-N curve. Thermal effect distance or escape distance Existing buildings New buildings pipeline. Site interaction distance FIGURE 7 - SITE-SPECIFIC PIPELINE INTERACTION DISTANCE FOR SOCIETAL RISK 6.9 The site-specific F-N data should be compared to the IGEM/TD/1 Edition 5 F-N criterion envelope. As the IGE TD/1 criterion envelope relates to a 1.6 km length of pipeline, the site-specific F-N criterion is obtained by factoring risk values by a value equal to 1.6 km divided by the site interaction distance. 21

29 6.10 The IGEM/TD/1 Edition 5 F-N criterion envelope represents broadly acceptable risk levels for pipeline routing. If the calculated site-specific F-N curve (which on a log F against log N plot would be expected to be generally rectangular in shape) is below the IGEM/TD/1 Edition 5 F-N criterion envelope, the risk levels to the adjacent population are considered broadly acceptable. An ALARP check may also be carried out as confirmation. If the site-specific F-N curve is far from the expected generally rectangular shape and cannot be approximately contained by a rectangle based on a point on the envelope, or approaches close to, or exceeds, the F-N criterion envelope, an ALARP demonstration will be required. Further mitigation may then be required to reduce risks to acceptable levels or the proposed development may be deemed unacceptable Where the methodology used in the assessment differs from that described in this Standard, and cannot be demonstrated to be conservative, an additional check should be made by assessing a relevant range of code compliant pipelines using the same methodology to locate the approximate position of the F-N envelope. 22